应用Squeeze算法实现地震数据高效压缩  

Applying Squeeze algorithm to achieve efficient compression of seismic data

作　　者：高潇 张伟 徐锦承 杨辉 邵理阳 潘权 宋章启 GAO Xiao;ZHANG Wei;XU Jincheng;YANG Hui;SHAO Liyang;PAN Quan;SONG Zhangqi(School of Earth and Space Sciences,the University of Science and Technology of China,Hefei 230000,China;Southern University of Science and Technology,Department of Earth and Space Sciences,Shenzhen 518055,China;Southern University of Science and Technology,Department of Electrical and Electronic Engineering,Shenzhen 518055,China;Southern University of Science and Technology,School of Innovation and Entrepreneurship,Shenzhen 518055,China)

机构地区：[1]中国科学技术大学地球和空间科学学院,合肥230000 [2]南方科技大学地球与空间科学系,深圳518055 [3]南方科技大学电子与电气工程系,深圳518055 [4]南方科技大学创新创业学院,深圳518055

出　　处：《物探化探计算技术》2020年第5期582-595,共14页Computing Techniques For Geophysical and Geochemical Exploration

基　　金：国家自然科学基金(41574107)。

摘　　要：为了提高海量地震数据的存储和处理能力,需要采用有损压缩方法对地震数据进行高效压缩。目前常用的地震有损压缩方法主要利用了地震数据的稀疏性来实现压缩,代码实现较为复杂,实际应用需要使用者对相应的压缩方法原理具有比较好地了解。另一种压缩实现原理是利用地震数据相邻道之间的关系实现压缩,该类方法原理简单,易于理解。这里将近年最新发展的Squeeze算法应用到地震数据压缩问题中,该算法通过使用多维和多层预测模型,基于邻近点信息预测当前点波场值,然后对实际波场值和预测值之间的差异,进行自适应量化编码和熵编码实现数据的有效压缩。在地震波数值模拟算法中实现了Squeeze压缩方法,并通过二维SEG/EAGE盐丘模型合成道集数据、实际三维道集数据以及三维Marmousi正演波场数据测试Squeeze压缩效果,结果表明,Squeeze算法对于地震数据具有比较好的压缩性能,同时代码实现和使用十分方便。In order to improve the storage and processing capacity of massive seismic data,it's necessary to use lossy compression methods to compress seismic data efficiently.The commonly used methods are mainly employ the seismic data sparsity to achieve compression,and the code implementations are also complicated,which requires the users to have a better understanding of the corresponding methods.Another compression implementation principle is to use the information of adjacent points of the seismic data to achieve compression.And the methods based on this are easy to be understood and applied.In this paper,we implement the newly developed algorithm,the Squeeze compression algorithm in seismic data compression.It employs a multi-layer and multi-dimensional prediction model for the relationship between adjacent points,and then performs adaptive quantization encoding for the difference between the real value and predicted value,and entropy encoding.We have implemented the Squeeze compression in the seismic forward modeling.And through the compression tests of 2D SEG-EAGE salt model post-stack data,real 3D seismic gather data,and the wavefield data of 3D Marmousi forward modeling,we show that the Squeeze algorithm has a good compression performance on seismic data,and its implementation and application are quite convenient.

关 键 词：地震数据 有损压缩 DCT Dreamlet SQUEEZE 地震波场 

分 类 号：P631.4[天文地球—地质矿产勘探]